Main point: Knowledge graphs based research gives overview on vague ideas,assists knowledge discovery, and is a strong collaboration tool.
Currently I am researching for at proposal on a project within machine learning and type theory. The idea, however, is not entirely condensed and figuring out where to pick the low hanging fruits is difficult. My process has often been that I create different documents that represent different aspects of the idea. In the end I eventually get lost in my own production to move on to more pressing tasks to forget about the project for weeks at the time.
This represents a class of ideas are also some that I get back to infrequently. I often revisit them after a long walk where I got some realization. Obviously, at that point I do not have the head space to put the new insight in context and rework old material. So it'll often just get chucked in there for later review. When I finally reach that point where I cam do a later review, the task might still be daunting. Even worse, one might have miscalculated the actual effort to clean up. In either direction. Sometimes I thought I had much more material than was actually the case.
This is a part of the create process and should not be discourages. But I definitely think there is room for tooling around it. This is why we are creating Spor Research. It is a tool that solves the overview and context problem. In particular we remove the meat of the research to leave on the skeletal structures. It is then significantly easier to figure out where to add in new insights, where to put a weekends worth of effort, and to serendipitously discover new connections.
First and foremost we integrate the tool with seconds brains, Zettelkasten, and linked notes. This is based on the assumption that research starts with what we know ourselves.
Secondly we integrate with public sources such as Wikipedia, arXiv (my hood), and business databases. This allows to explorer the context in-app. We have the idea that we need to remove barriers on thinking up search queries. So if we can just remove a fraction of dead-end searches we see that as an impact.
Lastly, we also know that research is not carried out individually. We also know that it is a big ask to make colleagues read and provide feedback on complicated texts. For these tasks we work on ways to share a research overview and also provide information on what participant of the teams has context and specific knowledge on each part.
We share updates on the progress on our LinkedIn Page.